TROPICS Vol. 16 (1) Issued January 31, 2007 Fruit visitation patterns of small mammals on the forest floor in a tropical seasonal forest of Thailand 1 2, 3 1 3 3 3 Shunsuke SUZUKI , Shumpei KITAMURA , Masahiro KON , Pilai POONSWAD , Phitaya CHUAILUA , Kamol PLONGMAI , Takakazu 2, 4 1 5 6 YUMOTO , Naohiko NOMA , Tamaki MARUHASHI and Prawat WOHANDEE 1 School of Environmental Science, The University of Shiga Prefecture, Hikone, 522−8533, Japan 2 Center for Ecological Research, Kyoto University, Kamitanakami-Hirano, Otsu, 520−2113, Japan 3 Thailand Hornbill Project, Department of Microbiology, Faculty of Science, Mahidol University, Bangkok 10400, Thailand 4 Research Institute of Humanity and Nature, 457-7, Motoyama, Kamigamo, Kita-ku, Kyoto, 602−0878, Japan 5 Department of Human and Culture, Musashi University, Nerima, Tokyo 176−8534, Japan 6 National Park, Wildlife and Plant Conservation Department, Bangkok 10900, Thailand Correspondence address: Shunsuke SUZUKI Tel: +81−749−28−8311, Fax: +81−749−28−8477, E-mail: [email protected] ABSTRACT The fruit visitation patterns of small birds, (Harrison, 1962; Miura et al. 1997), including mammals were investigated by camera trappings small mammals (Robinson et al. 1995; Wu et al. 1996; on the forest floor in a tropical seasonal forest of Yasuda et al. 2003; Pardini, 2004). The species richness Thailand. A total of 3,165 visits were recorded for and abundance of small mammals in tropical forests can seven small mammal species. The four Muridae be attributed to affluent food resources, especially the species, Rattus remotus, Niviventer fulvescens, abundant fruit crops (Fleming, 1973; August, 1983; Wells Leopoldamys sabanus and Maxomys surifer , all et al. 2004). Terrestrial small mammals are considered of which were nocturnal, were almost completely to have effects on tropical forest ecosystems as seed temporally segregated from the tree shrew, Tupaia and seedling predators and seed dispersers (Adler and belangeri , and the two squirrels, Callosciurus Kestell, 1998; Asquith et al. 1999; Guariguata et al. 2000). finlaysonii and Menetes berdmorei , which were In recent years, the abundance of large animal species diurnal or crepuscular. We suggest that the has declined and communities have been altered due to temporal segregation reduced the interference habitat loss or fragmentation in tropical forests (Soule et competition between the four Muridae species and al. 1992; Laidlaw, 2000; Sodhi et al. 2004). Small mammal the tree shrew or squirrel for fruits on the forest species, however, are adaptable enough to inhabit in floor. In addition, the visitation patterns for fruit various and patchy environments and are more robust species differed among the four Muridae species in dealing with habitat changes. In order to predict the and between the tree shrew and the two squirrels, fate of fragmented or disturbed forests, it is important to suggesting that the variation in the visitation understand the coexistence patterns of small mammal patterns for fruit species helped to facilitate assemblages in relation to their roles in tropical forest coexistence among these species. In contrast, ecosystems (Wells et al. 2004). the two squirrels were similar in their visitation The coexistence of sympatric species can be patterns, both temporally and in their choice of facilitated by the differentiation in niche dimensions, fruit species. such as habitat, food and time (Schoener, 1974; , M Closkey, 1976). There have been several studies on Key words: coexistence, crepuscular, diurnal, microhabitat utilization and segregation of small mammal frugivore, Khao Yai National Park, nocturnal, small assemblages in tropical forests (Shanker, 2001; Wells mammal, temporal segregation et al. 2004). While there have been some studies on food resource partitioning in rodents in the desert (e.g. Brown et al. 1979; Brown, 1989), there have been very INTRODUCTION few on small mammals in tropical forests (Emmons, 1980; Tropical forests are enormously diverse and complex Smythe, 1986). Moreover, little is known about temporal habitats and harbor many species of mammals and differentiation of food resource utilization. This is 18 Shunsuke SUZUKI, Shumpei KITAMURA, Masahiro KON, Pilai POONSWAD, Phitaya CHUAILUA, Kamol PLONGMAI, Takakazu YUMOTO, Naohiko NOMA, Tamaki MARUHASHI and Prawat WOHANDEE partly due to the difficulty in observing the often cryptic Park in Thailand. We obtained a large number of pictures behavior of small mammals. Temporal differentiation of small mammals visiting the fruit bait on the forest in food resource utilization can be recorded by various floor. In this paper, we examine the differentiation in the methods, including radio telemetry, consecutive fruit visitation patterns among small mammal species trapping, direct observation and camera trapping. In that utilize fruits on the forest floor, and compare the recent years, remarkable improvements have been made distribution of these patterns between small mammal for various wildlife studies through the use of camera species, both temporally and in terms of fruit species. trapping (Griffiths and van Schaik, 1993; Miura et al. 1997; Blanchong and Smale, 2000; Jayasekara et al. , 2003; O Brien et al. 2003; Kawanishi and Sunquist, 2004; MATERIALS AND METHODS Yasuda et al. 2005). Camera trapping can provide precise Study site time data for visitation patterns of animals, which can be The study was carried out in Khao Yai National Park quantitatively analyzed for visitation time and resource (hereafter KY; Fig. 1) in lower northeastern Thailand. , , utilization patterns (Yasuda et al. 2005). The park ranges from 14˚05 N to 14˚15 N and from 101˚ , , Through the use of camera trappings, the 05 E to 101˚ 50 E in the Dongruk mountain range, and 2 relationships between fruits and frugivores on the forest covers an area of 2,168 km . Its elevation ranges from floor of a tropical forest were studied in Khao Yai National 250 to 1,351 m. The main study site was adjacent to the 2 headquarters of KY and covered about 70 km . The area ranges from 600 to 800 m in altitude and is covered by moist evergreen forest. The annual mean rainfall is 2,360 mm (1993−2002); the wet season usually occurs from April to October and the dry season is from November to March (Kitamura et al. 2004). The monthly mean temperature ranges from 21˚C (December and January) to 32˚C (April and May). In the study area, sunrise is between 0545 h and 0643 h throughout a year and sunset is between 1743 h and 1847 h. According to the standard time near KY, we �������� divided a day into the following three time zones: daytime ������������� from 0700 h to 1700 h; nighttime from 1900 h to 0500 h; and crepuscular time from 0500 h to 0700 h and 1700 h to 1900 h. ������� Although ripe fruits (e.g. Ficus spp.) are available year-round (Poonswad et al. 1998), fruit diversity and abundance are relatively high in the rainy season and reach a trough at the beginning of the dry season (Bartlett 2003). Trees reach 45 m in height; the density of trees over 10 cm in diameter at breast height (DBH) is 371 ha−1, with a basal area of 32 m2ha−1. The dominant plant families in the forest include the Lauraceae, Cornaceae, Euphorbiaceae, Meliaceae, Dipterocarpaceae, and Annonaceae. Kitamura et al. (2005) provided a detailed botanical inventory. Photographing by automatic camera systems We conducted surveys on the fruit visitation patterns of small mammals from July 2000 to June 2002 using automatic camera systems. We used two kinds of systems, both of which consist of a far-infrared sensor, Fig. 1. Location of Khao Yai National Park motor-driven compact camera with a built-in flash and Fruit visitation patterns of small mammals in Thailand 19 a data pack that stamps each picture with the date and based only on the data for the fruit species that had been time of exposure. The system used until June 2001 was visited by both animal species. developed by Miura et al. (1997). From April 2001, we used a Sensor Camera“ FIELDNOTE” (MARIF Co. Ltd, Similarity in the visitation pattern for fruit species Japan; Yasuda, 2004). The cameras were checked at an Similarity in the rank order for the number of visits per interval of 1 to 3 days, and films (ISO 100 or ISO 400, 36 day to a fruit species was measured for each pair of , exposures) and batteries (CR123A lithium battery) were animal species using Spearman s rank correlation. If renewed if necessary. both of two given animal species have the same rank We used the fruits of 186 individuals of 67 plant order, then the rank correlation coefficient culminates in species as baits. Five to 50 fruits that had been collected 1; if animals have a completely inverse rank order, it is −1; from the ground or trees were set on the ground under and if independent of each other, it is 0. the source tree as bait, and a camera was placed about 2 m apart from the bait fruits. When the bait fruits were consumed by animals or damaged by insects or microbes, RESULTS they were replaced with new ones. The photographing Forty-eight animal species were identified from 11,079 duration varied from 5 to 90 consecutive days according pictures. Among them, eight small mammal species to the period when fruits were available from each were included, but one species (Crocidura horsfildi) was individual tree. Based on the pictures, we recorded excluded from further analyses due to the small sample when and what small mammals visited the bait fruits. size (< 10 visits). The remaining seven species were We followed Corbet and Hill (1992) for the taxonomic recognized from 5,419 pictures. From the pictures of nomenclature of small mammals. small mammal species, 3,165 (58.4%) were regarded as Because the cameras were triggered every 10 to 15 independent visits and used for further analyses.